Free Radical Biology & Medicine, Vol. 27, Nos. 11/12, pp.

13571366, 1999
Copyright 1999 Elsevier Science Inc.
Printed in the USA. All rights reserved
0891-5849/99/$see front matter

PII S0891-5849(99)00179-3

Original Contribution
CELLULAR ANTIOXIDANT AND PRO-OXIDANT ACTIONS OF
NITRIC OXIDE
MAHESH S. JOSHI,* JULIE L. PONTHIER,

and JACK

R. LANCASTER, JR.,

*Department of Surgery, Louisiana State University Medical Center, New Orleans, LA, USA; Department of Physiology, and

Department of Medicine, Louisiana State University Medical Center, New Orleans, LA, USA
(Received 3 December 1998; Revised 2 June 1999; Accepted 24 August 1999)

AbstractWe describe a biphasic action of nitric oxide (NO) in its effects on oxidative killing of isolated cells: low
concentrations protect against oxidative killing, while higher doses enhance killing, and these two effects occur by
distinct mechanisms. While low doses of NO (from (Z)-1-[N-(3-ammonio propyl)-N-(n-propyl)-amino]-diazen-1-ium1,22 diolate [PAPA/NO] or S-nitroso-N-acetyl-L-penicillamine [SNAP] prevent killing of rat hepatocytes by
t-butylhydroperoxide (tBH), further increasing doses result in increased killing. Similar effects occur with rat hepatoma
cells treated with PAPA/NO and tBH or H2O2. Increased killing with higher concentrations of NO donor is due to both
NO and tBH, because NO donor alone is without effect. Glutathione (GSH) is not involved in either of these actions.
Based on measurements of thiobarbituric acidreactive substances (TBARS) and effects of lipid radical scavenger
(DPPD) and deferoxamine, the protective effect, but not the enhancing effect, involves peroxidative chemistry. Fructose
has no effect on tBH killing alone but provides substantial protection against killing by higher concentrations of NO plus
tBH, suggesting that the enhancing effect involves mitochondrial dysfunction. Hepatocytes, when stimulated to produce
NO endogenously, become resistant to tBH killing, indicative of the presence of an NO-triggered antioxidant defensive
mechanism. The finding that the protective effects of low concentrations of NO and the harmful effects of high
concentrations of NO are fundamentally different in nature suggest that therapeutic interventions could be designed,
which selectively prevent its pro-oxidant activity at high concentrations, thus converting NO from a Janus-faced
modulator of oxidant injury into a pure protectant. 1999 Elsevier Science Inc.
KeywordsNitric oxide, Pro-oxidant, Antioxidant, Mitochondria, Radicals, t-Butylhydroperoxide, Hydrogen peroxide,
Stress proteins, Free radical

INTRODUCTION

[13]. This may cause genotoxicity and cytotoxicity due

to damage to key proteins and DNA. Therefore, NO
plays multifaceted roles depending on the conditions and
levels of its production and type of oxidative stress that
is induced. Abnormal production of NO has been linked
to a number of clinically important conditions.
NO is produced endogenously by several types of
mammalian cells including hepatocytes in response to
inflammatory stimuli, as a result of upregulation of the
inducible form of NO synthase (iNOS). NO is also
generated at low levels by another isoform called constitutive NOS (cNOS). These isoforms catalyze the
mixed-function oxidation of a guanidino nitrogen of Larginine to form NO and L-citrulline. However, the level
of NO production and its functional role can vary between cell types. The interactions between NO and reactive oxygen species (ROS) have assumed greater im-

Nitric oxide (NO, nitrogen monoxide), produced by a

variety of cell types, plays diverse roles as an immunomodulator, neurotransmitter, and vasodilator. By virtue
of its free radical nature, NO can participate in a wide
spectrum of biologically important reactions. The biological actions of NO are mediated by its reactions with
oxygen species and transition metals. The cellular targets
of the products of these reactions include iron-containing
proteins and complexes, low and high molecular weight
thiols, amine-containing targets such as nucleic acids,
phenolic groups such as tyrosine, and cellular radicals
Address correspondence to: Jack R. Lancaster, Jr., Department of
Physiology, Louisiana State University Medical Center, 1901 Perdido
Street, New Orleans, LA 70112, USA; Tel: (504) 568-8096; Fax: (504)
568-6158; E-Mail: jlanca@lsumc.edu.
1357

1358

M. S. JOSHI et al.

portance due to the simultaneous generation of both in

various pathophysiological conditions. Such interactions
follow different pathways resulting either in protection
or production of more toxic end products. NO has been
shown to be cytoprotective through its reaction with lipid
radicals [4] as well as with iron [5,6]. Wink et al. have
shown that either H2O2 [7] or t-Butyl hydroperoxide
(tBH) [8] mediated cytotoxicity can be attenuated by the
presence of NO releasing diazen-1-ium-1,2-diolates
(NONOates). By pretreating the cells with low concentrations of NO, we have previously demonstrated that
cells can mount defensive responses against ROS mediated injury [9]. It has also been proposed that NO can
enhance ROS toxicity due to its rapid reaction with

superoxide (O
2 ) to form peroxynitrite (ONOO ) [10].
Recently, formation of nitrosoperoxocarbonate anion
(ONOOCO
2 ) as a result of rapid reaction between
ONOO and CO2 has been demonstrated [11]. The nitrosoperoxocarbonate, thus formed, has been proposed to
act as an oxidant and nitrating species in biological
systems. The formation of ONOO thus could dramatically increase the toxicity associated with the presence of
either O
2 or NO alone. The majority of xenobiotic
oxidants are not biologically active but are converted to
reactive species by the target cell itself. tBH is one such
oxidant, chosen for the present investigation of the role
of NO in oxidative stress. In this study, we have attempted to clarify the conditions under which NO can be
protective or damaging in the presence of tBH in cells.
We show that lesser amounts of NO are protective
against killing by tBH and also by H2O2 while higher
concentrations enhance killing. The killing with tBH and
greater amounts of NO is fundamentally different in
mechanism than killing by tBH alone. We also show that
pretreatment of hepatocytes with cytokines, to generate
NO, provides protection against tBH toxicity.
MATERIALS AND METHODS

was a generous gift from David Wink (National Institutes of Health, Bethesda, MD, USA). Williams media
E, trypan blue, penicillin, streptomycin, L-glutamine
and N-[2-hydroxyethyl]piperazine-N-[2-ethanesulfonic
acid] HEPES were purchased from Gibco Laboratories.
Fetal calf serum was from Hyclone Laboratories (Logan,
UT, USA). Percoll was obtained from Pharmacia (Piscataway, NJ, USA) and collagenase from Worthington
Biochemical Corp. SNAP (S-nitroso-N-acetyl-L-penicillamine) was synthesized every 2 months as described
[12], stored frozen as a solid aliquot in the dark and
checked for stoichiometric S-nitrosothiol content by the
method of Saville [13].
Isolation, cell culture, and treatment of hepatocytes
Hepatocytes were isolated from male Sprague Dawley
rats (175 g) using a modification of in situ collagenase
perfusion technique as described previously [14]. The
cells thus obtained were purified further by Percoll gradient centrifugation to remove nonparenchymal cells.
The cells were cultured overnight at 37C in an atmosphere of 5% CO2 95% air in Williams medium E
containing 1-M insulin, 2-mM glutamine, 105 U/l penicillin, 100-mg/l streptomycin, and 10% low endotoxin
fetal calf serum in 12-well plates at a density of 120,000
cells per well. The cultures were incubated in Williams
medium E (without serum) with additions indicated in
the text for 5 h at 37C. At the end of treatment the cell
viability was determined by assaying for lactate dehydrogenase (LDH) in the culture supernatant [15].
For the pretreatment with cytokine mixture (CM),
overnight cultures were exposed to Williams medium E
(5% serum) containing 100-U/ml rat Interferon (IFN-),
500-U/ml murine tumor necrosis factor- (TNF-),
100-U/ml human recombinant interleukin-1 (IL-1),
and 10-g/ml lipopolysaccharide (LPS) for 24 h at 37C
in the presence or absence of iNOS inhibitor, 0.5-mM
ng-monomethyl-L-arginine.

Chemicals
tBH, menadione, glutathione reductase, deferoxamine
mesylate, butylated hydroxytoluene (BHT), NADPH,
DPPD (N,N-diphenyl-p-phenylenediamine), 1-Butanol,
2-thiobarbituric acid, NMMA (NG-monomethyl-L-arginine), lactate dehydrogenase (LDH) assay kit were from
Sigma (St. Louis, MO, USA). 2-Vinylpyridine and malonaldehyde bis(dimethyl acetal) were from Aldrich
(Milwaukee, WI, USA). GSSG and insulin were from
Boehringer Mannheim (Mannheim, Germany). 5,5-Dithiobis(2-nitrobenzoic acid) (DTNB) was from Fluka
(Ronkonkoma, NY, USA), and PAPA/NO ((Z)-1-[N(3-ammoniopropyl)-N-(n-propyl)-amino]-diazen-1-ium1,2-diolate) and PPD (N-propyl-1,3-propanediamine)

Culture and treatment of FTO2B cells

The rat hepatoma cell line FTO2B was grown at 37C
in a 5% CO2 atmosphere in Dulbeccos modified Eagles
medium/Hams F12 medium (1:1) supplemented with
10% heat-inactivated bovine calf serum, 2-mM glutamine, 50-IU/ml penicillin-streptomycin, and 0.2% sodium bicarbonate. FTO2B cells (2 105/ml) were plated
in 12-well plates ranging from passage number 5 to 6,
and allowed to adhere overnight before an experiment.
Cells were treated with a range of PAPA/NO doses from
1 M to 5 mM in addition to either 500-M tBH for 3 h
or 1-mM H2O2 for 2 h. Cell viability was determined by
assaying the culture supernatant for LDH.

Antioxidant and pro-oxidant actions of NO

1359

Oxidized glutathione (GSSG) assay

The method as described by Luperchio et al. [16] was
followed with some modifications. The culture supernatant was acidified with 10-mM HCl and frozen at 80C.
For the analysis of GSSG content, 70-l acidified supernatant was derivatized by addition of 1.5-l of 2-vinylpyridine (97% pure), vortexed vigorously and incubated at room temperature for 60 min and an aliquot was
used for the assay.

TBARS assay for lipid peroxidation

Thiobarbituric acidreactive substances (TBARS)
were measured by modification of the method of Buege
and Aust [17]. Briefly, to 0.8 ml of culture supernatant
was added 80-l 50% TCA, vortexed and centrifuged at
2000 rpm for 5 min. To 0.6 ml of supernatant was added
75 l of 90-mM butylated hydroxytoluene (BHT), 0.6 ml
of 0.2-M phosphoric acid and 75 l of 0.11-M 2-thiobarbituric acid, vortexed and boiled in a waterbath for 40
min. The reaction was stopped by placing the tubes in
ice-cold water. Then 0.15 ml of saturated NaCl, and 1.5
ml of 1-butanol were added, vortexed, and centrifuged at
2000 rpm for 5 min. The fluorescence of the organic
layer was read at excitation 515 nm and emission 553
nm. Malonaldehyde bis(dimethyl acetal) was used as a
standard.
RESULTS

Overnight cultures of rat hepatocytes were treated

with 175 M tBH for 5 h in the presence or absence of
increasing concentrations of NO donors SNAP or PAPA/
NO. PAPA/NO has a half-life of 77 min at 22C [18] and
hence is expected to liberate virtually all NO during the
5-h cell treatment. During the initial studies the cell
viability was determined by both crystal violet assay and
LDH assay. The results obtained by the two assays were
virtually identical and the more convenient LDH assay
was used for all subsequent studies. Figure 1 shows that
175-M tBH alone results in greater than 80% cell
killing whereas addition of low amounts (20 200 M)
of either of these two NO donors significantly protects
against tBH induced cellular toxicity. PAPA/NO acts as
a slightly better protectant (Fig. 1B) than SNAP (Fig.
1A). Increasing concentration of either donor results in
significant increase in cell killing. The treatment of cells
with SNAP or PAPA/NO alone (in the absence of tBH)
does not cause toxicity to the cells even at a concentration of 0.5 mM (open circles). This shows that the
increased killing with increasing donor requires the simultaneous presence of NO donor and tBH. In effect, the

Fig. 1. Biphasic effect of NO donor on tBH killing of hepatocytes.

Overnight cultures of rat hepatocytes were treated with 175-M tBH in
the presence or absence of increasing doses of either SNAP (A) or
PAPA/NO (B). After a 5-h treatment, cell killing was assessed by the
release of LDH into the supernatant. tBH plus SNAP or PAPA/
NO; O SNAP or PAPA/NO alone. Results are means SD of
samples in triplicates.

presence of tBH converts increasing NO donor from

being without effect to a toxic cell insult.
Figure 2 demonstrates the effects of PAPA/NO on
killing of the FTO2B rat hepatoma cell line by 500-M
tBH (Fig. 2A) or 1-mM H2O2 (Fig. 2B). Lower concentrations of NO donor protect against killing by either
oxidant and killing increases with higher donor concentration, similar to the effects above with primary hepatocytes. Increased killing with NO donor is due to an
interaction between both NO and oxidant, because killing does not increase with NO donor alone in the absence
of oxidant. These results show that both the antioxidant
effect of lower NO and the pro-oxidant effect of higher
NO are also present in a cultured cell line and that it also
can occur with H2O2 as oxidant.

1360

M. S. JOSHI et al.

Fig. 2. Biphasic effect of NO donor on tBH or H2O2 killing of

hepatoma cells. Rat hepatoma cells were treated with either 500-M
tBH for 3 h (A), or 1 mM H2O2 for 2 h (B) in the presence or absence
of increasing doses of PAPA/NO. Killing was assessed by LDH release
into the supernatant. PAPA/NO plus tBH (A) or H2O2 (B); O
PAPA/NO alone. Results are means SD of samples in triplicates.

To determine whether the protection observed with

low concentrations of NO donors or potentiation of killing with higher NO donor concentrations are due to NO
released from PAPA/NO, the decomposition products of
NO donors (namely degraded PAPA/NO, nitrite, or
N-propyl-1,3-propane diamine [PPD]) were tested. Nitrite, in particular, has been found to be toxic to Chinese
hamster V79 cells in the presence of tBH [8]. To test this
possibility, cells were treated with tBH in the presence of
20 M each of nitrite, degraded PAPA/NO (dPAPA/
NO), or PPD. As depicted in Fig. 3A, none of these
additions provide any protection from tBH alone, showing that the observed protection with low PAPA/NO
concentration is due to NO. As shown in Fig. 3B, when
cells are protected from tBH by the presence of low
PAPA/NO concentration, addition of degraded PAPA/

Fig. 3. Biphasic effect of NO donor is due to NO liberation. Overnight

cultures of rat hepatocytes were treated with 175-M tBH with various
additions for 5 h and the cell killing was assessed by the release of LDH
into the supernatant. Degraded PAPA/NO was prepared by keeping a
10 mM solution of PAPA/NO in a 37C incubator overnight. Results
are means SD of samples in triplicates.

NO, PPD, or nitrite does not increase killing, whereas

addition of more PAPA/NO does. Note that this higher
PAPA/NO without tBH is not toxic (Fig. 1). This shows
that the increased killing with higher PAPA/NO concentration is due to NO. Thus, low NO antagonizes killing,
and higher NO acts in combination with tBH to increase
killing.
We were interested in delineating the major characteristics of these antioxidant/pro-oxidant actions of NO,
in particular, whether the effects of increasing NO are

Antioxidant and pro-oxidant actions of NO

due to a simple reversal of the protective effect of low

NO or initiation of a toxic injury from NO/tBH interaction which is different from killing by tBH alone.
The metabolism of tBH by hepatocytes results in
oxidation of GSH, an important cellular antioxidant. This
loss is accounted for by the accumulation of oxidized
glutathione (GSSG) in the culture medium [19]. The
protection observed against tBH killing with low NO
could involve the inhibition of intracellular GSH oxidation. Hence the levels of GSSG were measured in the
culture supernatant. Addition of either low or high concentrations of NO has no effect on the tBH initiated
appearance of GSSG (Fig. 4A). Thus, the protective
effects of low concentrations of NO are not due to
attenuation of glutathione oxidation, and likewise the
increased injury due to NO/tBH interaction does not
involve glutathione. We conclude this because there is no
change in glutathione oxidation upon addition of higher
concentrations of NO (50 M donor) (Fig. 4A)
whereas there is a dramatic increase in killing (Fig. 1).
The involvement of metal-catalyzed peroxidative injury (particularly lipid peroxidation) in the cytotoxicity
of tBH has been demonstrated previously [19 21]. The
protection afforded by low levels of NO could be the
result of inhibition of lipid peroxidation and the cytotoxicity resulting from the presence of high NO during tBH
treatment may involve lipid peroxidation. To test this
possibility, the extent of lipid peroxidation was measured
in the culture supernatant. Although not completely specific, the TBARS assay has been widely used to gauge
the extent of lipid peroxidation [2224]. As shown in
Fig. 4B, the presence of low as well as high concentrations of NO during tBH treatment completely prevents
the lipid peroxidation caused by tBH as determined by
TBARS assay. Thus, the protection by low concentrations of NO is associated with inhibition of lipid peroxidation and the cytotoxicity of high concentrations of NO
plus tBH is independent of membrane lipid peroxidation.
The TBARS assay, although widely used to measure the
lipid peroxidation, is not specific. Therefore, these results
obtained with TBARS assay were verified by using an
inhibitor of lipid peroxidation. DPPD, an antioxidant that
protects against tBH killing by scavenging lipid radicals
[19], although protective against tBH alone, fails to protect the cells from tBH plus higher NO toxicity (Fig. 5A).
Similarly, deferoxamine, an iron chelator, also fails to
provide protection at high NO levels (Fig. 5B) implying
that, unlike tBH killing, the killing due to tBH plus high
NO is not mediated by iron.
It has been reported previously that mitochondrial
injury can be involved in tBH-induced cell killing and
that fructose provides protection against this injury [25,
26]. Fructose is an efficient substrate for glycolytic ATP
generation by liver cells and has been shown to protect

1361

Fig. 4. Role of glutathione depletion and TBARS formation in low and

high PAPA/NO effects on tBH treatment. Overnight cultures of rat
hepatocytes were treated with 175-M tBH in presence and absence of
PAPA/NO for 5 h. The culture supernatant was assayed for (A) GSSG
content or (B) TBARS as described in Materials and Methods. Results
are means SD of samples in triplicates.

against injury attributable to ATP depletion [27]. In our

system, addition of 20-mM fructose affords dramatic
protection against high NO plus tBH killing (Fig. 5C) but
not killing with tBH alone. This result suggests that the
cytotoxicity by high concentrations of NO plus tBH may
involve bioenergetic dysfunction.
Relatively few studies have been done on effects of
endogenous NO on ROS injury using isolated cells in
culture [9,28 31]. This is important because biological
conditions may not be favorable for some chemical reactions of nitrogen oxides that occur in the test tube.

1362

M. S. JOSHI et al.

Fig. 5. Effect of lipid radical scavenger, iron chelator, or fructose on tBH plus low and high PAPA/NO mediated cell killing. Overnight
cultures of rat hepatocytes were treated with 175-M tBH in presence and absence of PAPA/NO for 5 h, in the presence of 1-M DPPD
(A), 20 mM deferoxamine (B), or 20 mM fructose (C). The cell killing was assessed by the release of LDH into the supernatant. Results
are means SD of samples in triplicates.

Therefore, we studied the effect of pretreatment of hepatocytes with endogenously generated NO against tBH
induced toxicity. Cell cultures were induced for NO
production by treatment with a mixture of inflammatory
mediators [9] in the presence or absence of NG-monomethyl-L-arginine (L-NMMA, an inhibitor of NOS) for
24 h and then media were replaced with 175-M tBH
plus L-NMMA for 3 h. The inclusion of L-NMMA
during the tBH exposure assures that the effects we are
detecting are due to the prior exposure of the cells to their
own NO synthesis (as differentiated from NO synthesis
that is ongoing during the tBH treatment). As depicted in
Fig. 6, cytokine pretreatment significantly protects
against tBH killing and this killing can be reversed by
L-NMMA, showing that protection by the inflammatory
mediator treatment is due to NO synthesis. Pretreatment
with L-NMMA alone (without inflammatory mediator
pretreatment) has no effect on tBH toxicity.

sible for cell death can vary depending on the experimental conditions, phenomena associated with oxidant
injury occur with tBH-treated cells including metal-catalyzed lipid peroxidation [19 21,34], glutathione deple-

DISCUSSION

NO has been shown to possess either antioxidant or

pro-oxidant properties [13,32,33]. The factors that dictate this protective versus damaging effect of NO on
oxidative injury remain poorly defined. We have addressed this issue by studying the effects of simultaneous
addition of NO donors and ROS-generating compounds
to cell cultures. We focussed on tBH as a model compound to generate oxidative stress because its oxidative
effects on hepatocytes have been rather extensively studied previously. Although the major phenomena respon-

Fig. 6. Previous synthesis of NO by hepatocytes protects them against

tBH killing. Overnight cultures of rat hepatocytes were pretreated with
media containing cytokine mixture (CM), CM NMMA, or NMMA
alone for 24 h. These cultures were then exposed to 175-M tBH
NMMA for 3 h and cell killing was assessed by the release of LDH into
the supernatant. Results are means SD of samples in triplicates.

Antioxidant and pro-oxidant actions of NO

tion [35,36], dysregulation of intracellular calcium [37,

38], and mitochondrial dysfunction [25,26,39 41].
tBH requires longer exposure times to cause appreciable cytotoxicity. Hence to study the effect of NO, we
need a continuous supply of NO. A class of compounds
called NONOates meets this requirement by spontaneously liberating NO at a controlled rate in solution [42].
We have found that addition of lower doses of NO as
PAPA/NO or SNAP to either isolated rat hepatocytes
(Fig. 1) or cultured rat hepatoma FTO2B cells (Fig. 2)
attenuates tBH toxicity whereas higher doses potentiates
it. This biphasic effect of NO is also exhibited by FTO2B
cells treated with H2O2 (Fig. 2). The increasing toxicity
in the presence of tBH plus increasing NO dose (Figs. 1
and 2) shows that this toxicity is due to an interaction
between NO and tBH, because increasing doses of NO
donors alone does not cause cytotoxicity. It can be argued that at higher concentrations NO does not potentiate tBH toxicity but instead no longer provides any
protection. In addition to mechanistic problems with this
possibility (why should increasing concentrations of a
protectant, in the absence of any other phenomenon,
reverse the effects of the lower concentrations already
present?), we have experimental evidence that this is not
the case. The effects of DPPD, deferoxamine and fructose demonstrate that the two effects (killing due to tBH
alone and to tBH plus higher concentrations of NO) must
be due to separate mechanisms. In addition, killing by
tBH alone is associated with increased TBARS formation whereas killing by higher NO/tBH is not. Thus,
while the killing mechanism induced by tBH alone is
prevented by low (and also high) concentrations of NO,
at higher concentrations there is a separate mechanism of
killing which occurs and it involves cellular toxicity
from the combination of tBH plus NO. Therefore, by
increasing the dose, NO switches from an antioxidant to
pro-oxidant. Wink et al. [8] have also seen PAPA/NO
protection against tBH toxicity in Chinese hamster lung
fibroblasts and shown that NO flux rate is an important
determinant of its actions, but they did not study effects
of higher doses of NO.
The question arises as to how the amount of NO
during these in vitro treatments compares to the amounts
encountered by cells in vivo. Using electrochemical detection, it has been found that the concentrations of NO
under nonpathological conditions (i.e., agonist stimulated endothelial cells) is in the nanomolar range,
whereas NO levels under conditions of oxidant injury
(middle cerebral artery occlusion) are in the micromolar
range [43]. Although it is difficult to speculate about the
steady-state NO concentration with SNAP because its
breakdown rate is influenced by a variety of uncontrolled
factors (such as metal-ion catalysis [44] and cellular
metabolism [45]), the diazeniumdiolates such as PA-

1363

PA/NO spontaneously break down liberating NO in a

well-described first order process [18]. We find that the
antioxidant/pro-oxidant switch occurs with PAPA/NO
concentrations in the 0.1 0.5-mM concentration range
(Figs. 1 and 2), which corresponds to steady-state NO
concentrations from PAPA/NO in the low micromolar
range [8]. This suggests that these results are relevant to
physiological conditions, i.e., low NO levels being protective and high levels generated during pathological
conditions being damaging.
There are several possible mechanisms through which
low NO levels may be providing protection. As a result
of the cellular metabolism of tBH in the presence of iron,
alkoxyl, and peroxyl radicals may be formed. These
radicals initiate formation of lipid radicals leading to
membrane lipid peroxidation and cell death [19 21,34].
Using phosphatidylcholine liposomes it has been demonstrated that NO can directly react with lipid alkoxyl
and peroxyl radicals and thus terminate lipid radical
chain propagation reactions [4]. Second, NO is widely
known to bind to iron to form iron-nitrosyl complexes
and prevent metal-catalyzed ROS formation [5], thus
potentially inhibiting the formation of alkoxyl and peroxyl radicals from tBH. NO has been shown to protect
neurons as well as rat hepatocytes against iron-induced
oxidative stress [46,47]. Gorbunov et al. [6] have demonstrated that in erythroleukemia cells, NO inhibits tBH
mediated oxidative damage by nitrosylation of heme and
non-heme iron. Finally, it is also possible that NO
quenches the effects of tBH by reacting with t-butylperoxyl radical, a reaction which is known to be fast [48].
tBH metabolism by hepatocytes results in concomitant loss of intracellular GSH levels (Fig. 4A). Inhibition
of tBH toxicity by deferoxamine or DPPD without depletion of GSH has been reported previously [19], indicating that the cell killing was independent of GSH
metabolism by the cells. We observed a similar phenomenon with respect to NO. The inhibition of tBH toxicity
by low NO had no effect on the tBH-induced increase in
GSSG appearance in the culture supernatant. In another
model of oxidative stress, NO inhibited cell killing as
well as loss of GSH depletion due to glucose oxidase
mediated cytotoxicity [49]. Kuo and Slivka [29] have
observed that depletion of intracellular GSH associated
with acetaminophen-mediated hepatocyte injury was accelerated in the presence of NMMA, thus indicating NO
may inhibit the acetaminophen injury by modulating
GSH levels. Thus, antioxidant effects of NO may occur
by multiple mechanisms, with one important factor being
the nature of the oxidative injury.
What is the mechanism of tBH plus high NO dependent cell killing? Based on the results with DPPD and
deferoxamine (Figs. 5A and 5B), and also measurements
of TBARS formation (Fig. 4B), the mechanism appears

1364

M. S. JOSHI et al.

not to involve metal-catalyzed peroxidative events. One

possibility is that tBH depletion of glutathione (Fig. 4)
increases cellular sensitivity to nitrosative injury due to
less thiol available as a nucleophilic interceptor of NO
auto-oxidation products [50]. The protection observed
with fructose (Fig. 5C) suggests the involvement of
mitochondrial injury because it has been shown previously that fructose acts as an efficient substrate for glycolytic ATP synthesis and, therefore, is capable of replenishment of decreased ATP levels under oxidative
stress conditions [26,27,39,51]. Indeed, it has been
shown previously that high concentrations of tBH can
cause mitochondrial injury and cell death via a mechanism(s) not involving lipid peroxidation [25,26,41]. Our
results suggest that this injury is potentiated by NO. The
mechanism of this potentiation is the basis of future
work.
Another aspect of the role of NO in oxidative stress is
its ability to upregulate defense mechanisms against
ROS insult. As depicted in Fig. 6, NO pretreatment
provides dramatic protection against tBH toxicity. We
have demonstrated previously NO-induced protective
gene upregulation and also cross-resistance to H2O2 toxicity [9] and upregulation of mRNA for heme oxygenase
in response to production of endogenous NO [14]. We
suggest that increased HO activity results in decreased
heme and increased nonheme iron in the cell, which in
turn results in increased ferritin which counteracts the
injurious oxidative effects of this loosely-bound iron
[52]. Once ferritin is upregulated in response to the
increased nonheme iron, this increased ferritin provides
resistance to further oxidative injury. Therefore, cells
appear to mount a general protective response against
different kinds of cytotoxic agents. There is emerging
evidence that HO-induced ferritin upregulation may be
an important general mechanism of defense against oxidative stress [30,5355]. In addition to heme oxygenase
(hsp32) and ferritin, NO also upregulates hsp70 [31,56],
which may also afford protection.
There is evidence that the effects of NO on tissue
oxidative injury are similar in pattern to what we have
shown here with isolated cells, namely, that low amounts
of NO are protective against oxidative injury whereas
high concentrations are injurious [33,57,58]. Under these
conditions, it has generally been thought that therapeutic
interventions involving modulation of NO levels would
be difficult to accomplish because the amount of NO
would have to be maintained at a critical level, high
enough to be protective yet not high enough to be damaging. Our results suggest that these two mechanisms
(antioxidant and pro-oxidant) are in fact fundamentally
different in nature and, therefore, raise the possibility
that selective intervention might be accomplished
through antagonism of the high NO mechanism. For

example, we show here that fructose selectively prevents

the pro-oxidant effects of high NO while having no
effect on the protective, antioxidant effects of low NO
(Fig. 6B). Thus, administration of similar selective antagonists of this high NO process under clinically
relevant conditions (e.g., ischemia/reperfusion injury, inflammation, etc.) might convert NO into only a pure
protectant.
Finally, these results suggest an important consequence of the diffusional spread of NO away from a
single NO-producing cell or close collection of NOproducing cells [59]. Because the concentration of NO is
highest in the immediate vicinity of the source(s), declining in concentration with increasing distance, the
results here suggest that there may be a distinct spatial
boundary surrounding the NO source within which NO
will enhance oxidative injury and outside of which NO
will attenuate it. This effect may serve to isolate areas of
tissue oxidative injury to specific sites where NO is
produced.
In conclusion, data presented here show that low-level
NO acts as an antioxidant and higher-level NO as a
pro-oxidant when rat hepatocytes or cultured rat hepatoma cells are exposed to oxidative killing. The mechanism of low concentrations of NOs protection may
involve diminished metal-catalyzed lipid peroxidation
and the high concentration of NOs potentiation of oxidative stress may involve mitochondrial dysfunction. We
have also found that the pretreatment of cells with cytokines induces NO protection against tBH killing, possibly due to upregulation of protective proteins (including
heme oxygenase and ferritin) as we have demonstrated
previously for cell killing from high levels of NO and
from H2O2 [9].
Acknowledgements This work was supported by National Institutes
of Health Grant DK46935 to J.R.L. We thank Cornel Dobrescu and
John Bovastro, Jr. for excellent technical assistance.

REFERENCES
[1] Kerwin, J. F. J.; Lancaster, J. R., Jr.; Feldman, P. L. Nitric oxide:
a new paradigm for second messengers. J. Med. Chem. 38:4343
4362; 1995.
[2] Rubbo, H.; Darley-Usmar, V.; Freeman, B. A. Nitric oxide regulation of tissue free radical injury. Chem. Res. Toxicol. 9:809
820; 1996.
[3] Wink, D. A.; Hanbauer, I.; Grisham, M. B.; Laval, F.; Nims,
R. W.; Laval, J.; Cook, J.; Pacelli, R.; Liebmann, J.; Krishna, M.;
Ford, P. C.; Mitchell, J. B. Chemical biology of nitric oxide:
regulation and protective and toxic mechanisms. Curr. Topic Cell.
Regul. 34:159 187; 1996.
[4] Rubbo, H.; Radi, R.; Trujillo, M.; Telleri, R.; Kalyanaraman, B.;
Barnes, S.; Kirk, M.; Freeman, B. A. Nitric oxide regulation of
superoxide and peroxynitrite-dependent lipid peroxidation. Formation of novel nitrogen-containing oxidized lipid derivatives.
J. Biol. Chem. 269:26066 26075; 1994.
[5] Kanner, J.; Harel, S.; Granit, R. Nitric oxide as an antioxidant.
Arch. Biochem. Biophys. 289:130 136; 1991.

Antioxidant and pro-oxidant actions of NO

[6] Gorbunov, N. V.; Yalowich, J. C.; Gaddam, A.; Thampatty, P.;
Ritov, V. B.; Kisin, E. R.; Elsayed, N. M.; Kagan, V. E. Nitric
oxide prevents oxidative damage produced by tert-butyl hydroperoxide in erythroleukemia cells via nitrosylation of heme
and non-heme iron. Electron paramagnetic resonance evidence.
J. Biol. Chem. 272:12328 12341; 1997.
[7] Wink, D. A.; Hanbauer, I.; Krishna, M. C.; DeGraff, W.; Gamson,
J.; Mitchell, J. B. Nitric oxide protects against cellular damage
and cytotoxicity from reactive oxygen species. Proc. Natl. Acad.
Sci. USA 90:98139817; 1993.
[8] Wink, D. A.; Cook, J. A.; Krishna, M. C.; Hanbauer, I.; DeGraff,
W.; Gamson, J.; Mitchell, J. B. Nitric oxide protects against alkyl
peroxide-mediated cytotoxicity: further insights into the role nitric oxide plays in oxidative stress. Arch. Biochem. Biophys.
319:402 407; 1995.
[9] Kim, Y. M.; Bergonia, H.; Lancaster, J. R., Jr. Nitrogen oxideinduced autoprotection in isolated rat hepatocytes. FEBS Lett.
374:228 232; 1995.
[10] Beckman, J. S.; Beckman, T. W.; Chen, J.; Marshall, P. A.;
Freeman, B. A. Apparent hydroxyl radical production by peroxynitrite: implications for endothelial injury from nitric oxide
and superoxide. Proc. Natl. Acad. Sci. USA 87:1620 1624; 1990.
[11] Uppu, R. M.; Squadrito, G. L.; Pryor, W. A. Acceleration of
peroxynitrite oxidations by carbon dioxide. Arch. Biochem. Biophys. 327:335343; 1996.
[12] Field, L.; Dilts, R. V.; Ravichandran, R.; Lenhert, P. G.; Carnahan, G. E. An unusually stable thionitrite from N-acetyl-D,Lpenicillamine; X-ray crystal and molecular structure of
2-(acetylamino)-2-carboxy-1,1-dimethylethyl
thionitrite.
J. Chem. Soc. Chem. Commun. 249 250; 1978.
[13] Saville, B. A Scheme for the colorimetric determination of microgram amounts of thiols. Analyst 83:670 672; 1958.
[14] Kim, Y. M.; Bergonia, H. A.; Muller, C.; Pitt, B. R.; Watkins,
W. D.; Lancaster, J. R., Jr. Loss and degradation of enzymebound heme induced by cellular nitric oxide synthesis. J. Biol.
Chem. 270:5710 5713; 1995.
[15] Farber, J. L.; Young, E. E. Accelerated phospholipid degradation
in anoxic rat hepatocytes. Arch. Biochem. Biophys. 211:312320;
1981.
[16] Luperchio, S.; Tamir, S.; Tannenbaum, S. R. No-induced oxidative stress and glutathione metabolism in rodent and human cells.
Free Radic. Biol. Med. 21:513519; 1996.
[17] Buege, J. A.; Aust, S. D. Microsomal lipid peroxidation. Meth.
Enz. 52:302310; 1978.
[18] Hrabie, J. A.; Klose, J. R.; Wink, D. A.; Keefer, L. K. New nitric
oxide-releasing zwitterions derived from polyamines. J. Org.
Chem. 58:14721476; 1993.
[19] Masaki, N.; Kyle, M. E.; Farber, J. L. Tert-butyl hydroperoxide
kills cultured hepatocytes by peroxidizing membrane lipids. Arch.
Biochem. Biophys. 269:390 399; 1989.
[20] Sakaida, I.; Thomas, A. P.; Farber, J. L. Increases in cytosolic
calcium ion concentration can be dissociated from the killing of
cultured hepatocytes by tert-butyl hydroperoxide. J. Biol. Chem.
266:717722; 1991.
[21] Buc-Calderon, P.; Latour, I.; Roberfroid, M. Biochemical changes
in isolated hepatocytes exposed to tert-butyl hydroperoxide. Implications for its cytotoxicity. Cell Biol. Toxicol. 7:129 143;
1991.
[22] Hogberg, J.; Orrenius, S.; OBrien, P. J. Further studies on lipidperoxide formation in isolated hepatocytes. Eur. J. Biochem.
59:449 455; 1975.
[23] Hill, K. E.; Burk, R. F. Toxicity studies in isolated hepatocytes
from selenium-deficient rats and vitamin E-deficient rats. Toxicol.
Appl. Pharmacol. 72:3239; 1984.
[24] Carini, R.; Parola, M.; Dianzani, M. U.; Albano, E. Mitochondrial
damage and its role in causing hepatocyte injury during stimulation of lipid peroxidation by iron nitriloacetate. Arch. Biochem.
Biophys. 297:110 118; 1992.
[25] Masaki, N.; Kyle, M. E.; Serroni, A.; Farber, J. L. Mitochondrial
damage as a mechanism of cell injury in the killing of cultured

[26]

[27]

[28]

[29]
[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

[42]

[43]

1365

hepatocytes by tert-butyl hydroperoxide. Arch. Biochem. Biophys.

270:672 680; 1989.
Imberti, R.; Nieminen, A. L.; Herman, B.; Lemasters, J. J. Mitochondrial and glycolytic dysfunction in lethal injury to hepatocytes by t-butylhydroperoxide: protection by fructose, cyclosporin A and trifluoperazine. J. Pharmacol. Exp. Ther. 265:392
400; 1993.
Nieminen, A. L.; Saylor, A. K.; Herman, B.; Lemasters, J. J. ATP
depletion rather than mitochondrial depolarization mediates hepatocyte killing after metabolic inhibition. Am. J. Physiol. 267:
C67C74; 1994.
Nunoshiba, T.; deRojas-Walker, T.; Wishnok, J. S.; Tannenbaum,
S. R.; Demple, B. Activation by nitric oxide of an oxidative-stress
response that defends Escherichia coli against activated macrophages. Proc. Natl. Acad. Sci. USA 90:99939997; 1993.
Kuo, P. C.; Slivka, A. Nitric oxide decreases oxidant-mediated
hepatocyte injury. J. Surg. Res. 56:594 600; 1994.
Motterlini, R.; Foresti, R.; Intaglietta, M.; Winslow, R. M. NOmediated activation of heme oxygenase: endogenous cytoprotection against oxidative stress to endothelium. Am. J. Physiol.
270:H107H114; 1996.
Kim, Y. M.; de Vera, M. E.; Watkins, S. C.; Billiar, T. R. Nitric
oxide protects cultured rat hepatocytes from tumor necrosis factor-alpha-induced apoptosis by inducing heat shock protein 70
expression. J. Biol. Chem. 272:14021411; 1997.
Beckman, J. S.; Koppenol, W. H. Nitric oxide, superoxide, and
peroxynitrite: the good, the bad, and ugly. Am. J. Physiol. 271:
C1424 C1437; 1996.
Fehsel, K.; Kolb-Bachofen, V. Nitric oxide: cytotoxicity versus
cytoprotection - how, why, when, and where? Nitric Oxide 1:107
120; 1997.
Rush, G. F.; Gorski, J. R.; Ripple, M. G.; Sowinski, J.; Bugelski,
P.; Hewitt, W. R. Organic hydroperoxide-induced lipid peroxidation and cell death in isolated hepatocytes. Toxicol. Appl. Pharmacol. 78:473 483; 1985.
Shan, X.; Jones, D. P.; Hashmi, M.; Anders, M. W. Selective
depletion of mitochondrial glutathione concentrations by (R,S)3-hydroxy-4-pentenoate potentiates oxidative cell death. Chem.
Res. Toxicol. 6:75 81; 1993.
Nishida, K.; Ohta, Y.; Ishiguro, I. Modulating role of endogenous
reduced glutathione in tert-butyl hydroperoxide-induced cell injury in isolated rat hepatocytes. Free Radic. Biol. Med. 23:453
462; 1997.
Bellomo, G.; Jewell, S. A.; Thor, H.; Orrenius, S. Regulation of
intracellular calcium compartmentation: studies with isolated
hepatocytes and t-butyl hydroperoxide. Proc. Natl. Acad. Sci.
USA 79:6842 6846; 1982.
Nicotera, P.; McConkey, D.; Svensson, S. A.; Bellomo, G.; Orrenius, S. Correlation between cytosolic Ca2 concentration and
cytotoxicity in hepatocytes exposed to oxidative stress. Toxicology 52:55 63; 1988.
Nieminen, A. L.; Dawson, T. L.; Gores, G. J.; Kawanishi, T.;
Herman, B.; Lemasters, J. J. Protection by acidotic pH and fructose against lethal injury to rat hepatocytes from mitochondrial
inhibitors, ionophores, and oxidant chemicals. Biochem. Biophys.
Res. Commun. 167:600 606; 1990.
Shertzer, H. G.; Bannenberg, G. L.; Zhu, H.; Liu, R. M.; Moldeus,
P. The role of thiols in mitochondrial susceptibility to iron and
tert-butyl hydroperoxide-mediated toxicity in cultured mouse
hepatocytes. Chem. Res. Toxicol. 7:358 366; 1994.
Nieminen, A. L.; Byrne, A. M.; Herman, B.; Lemasters, J. J.
Mitochondrial permeability transition in hepatocytes induced by
t-BuOOH: NAD(P)H and reactive oxygen species. Am. J. Physiol.
272:C1286 C1294; 1997.
Maragos, C. M.; Morley, D.; Wink, D. A.; Dunams, T. M.;
Saavedra, J. E.; Hoffman, A.; Bove, A. A.; Isaac, L.; Hrabie, J. A.;
Keefer, L. K. Complexes of .NO with nucleophiles as agents for
the controlled biological release of nitric oxide. Vasorelaxant
effects. J. Med. Chem. 34:32423247; 1991.
Malinski, T.; Czuchajowski, L. Nitric oxide measurement by
electrochemical methods. In: Feelisch M; Stamler JS, eds. Meth-

1366

[44]
[45]
[46]

[47]

[48]
[49]
[50]

[51]

[52]
[53]

M. S. JOSHI et al.
ods in nitric oxide research. New York: John Wiley & Sons Ltd;
1996:319 339.
McAninly, J.; Williams, D. L. H.; Askew, S. C.; Butler, A. R.;
Russell, C. Metal ion catalysis in nitrosothiol (RSNO) decomposition. J. Chem. Soc. Chem. Commun. 1758 1759; 1993.
Kowaluk, E. A.; Fung, H. L. Spontaneous liberation of nitric
oxide cannot account for in vitro vascular relaxation by S-nitrosothiols. J. Pharmacol. Exp. Ther. 255:1256 1264; 1990.
Rauhala, P.; Mohanakumar, K. P.; Sziraki, I.; Lin, A. M.; Chiueh,
C. C. S-nitrosothiols and nitric oxide, but not sodium nitroprusside, protect nigrostriatal dopamine neurons against iron-induced
oxidative stress in vivo. Synapse 23:58 60; 1996.
Sergent, O.; Griffon, B.; Morel, I.; Chevanne, M.; Dubos, M. P.;
Cillard, P.; Cillard, J. Effect of nitric oxide on iron-mediated
oxidative stress in primary rat hepatocyte culture. Hepatology
25:122127; 1997.
Padmaja, S.; Huie, R. E. The reaction of nitric oxide with organic
peroxyl radicals. Biochem. Biophys. Res. Commun. 195:539 544;
1993.
Chang, J.; Rao, N. V.; Markewitz, B. A.; Hoidal, J. R.; Michael,
J. R. Nitric oxide donor prevents hydrogen peroxide-mediated
endothelial cell injury. Am. J. Physiol. 270:L931L940; 1996.
Wink, D. A.; Nims, R. W.; Darbyshire, J. F.; Christodoulou, D.;
Hanbauer, I.; Cox, G. W.; Laval, F.; Laval, J.; Cook, J. A.;
Krishna, M. C. Reaction kinetics for nitrosation of cysteine and
glutathione in aerobic nitric oxide solutions at neutral pH. Insights
into the fate and physiological effects of intermediates generated
in the NO/O2 reaction. Chem. Res. Toxicol. 7:519 525; 1994.
Di Monte, D.; Sandy, M. S.; Blank, L.; Smith, M. T. Fructose
prevents 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)induced ATP depletion and toxicity in isolated hepatocytes. Biochem. Biophys. Res. Commun. 153:734 740; 1988.
Kim, Y. M.; Bergonia, H. A.; Muller, C.; Pitt, B. R.; Watkins,
W. D.; Lancaster, J. R., Jr. Nitric oxide and intracellular heme.
Adv. Pharmacol. 34:277291; 1995.
Balla, G.; Jacob, H. S.; Balla, J.; Rosenberg, M.; Nath, K.; Apple,
F.; Eaton, J. W.; Vercellotti, G. M. Ferritin: a cytoprotective
antioxidant strategem of endothelium. J. Biol. Chem. 267:18148
18153; 1992.

[54] Lin, F.; Girotti, A. W. Hemin-enhanced resistance of human

leukemia cells to oxidative killing: antisense determination of
ferritin involvement. Arch. Biochem. Biophys. 352:5158; 1998.
[55] Oberle, S.; Schroder, H. Ferritin may mediate SIN-1-induced
protection against oxidative stress. Nitric Oxide 1:308 314; 1997.
[56] Malyshev, I. Y.; Malugin, A. V.; Golubeva, L. Y.; Zenina, T. A.;
Manukhina, E. B.; Mikoyan, V. D.; Vanin, A. F. Nitric oxide
donor induces HSP70 accumulation in the heart and in cultured
cells. FEBS Lett. 391:2123; 1996.
[57] Darley-Usmar, V.; Halliwell, B. Blood radicals: reactive nitrogen
species, reactive oxygen species, transition metal ions, and the
vascular system. Pharm. Res. 13:649 662; 1996.
[58] Whittle, B. J. Nitric oxidea mediator of inflammation or mucosal defence. Eur. J. Gastroenterol. Hepatol. 9:1026 1032;
1997.
[59] Lancaster, J. R., Jr. A tutorial on the diffusibility and reactivity of
free nitric oxide. Nitric Oxide 1:18 30; 1997.

ABBREVIATIONS

NOSnitric oxide synthase

tBHtert-butyl hydroperoxide
PAPA/NO(Z)-1-[N-(3-ammoniopropyl)-N-(n-propyl)-amino]-diazen-1-ium-1,2-diolate
SNAPS-nitroso-N-acetyl-L-penicillamine
NAPN-acetyl-L-penicillamine
GSH glutathione
GSSG oxidized glutathione
TBARSthiobarbituric acid reactive substances
DPPDN,N-diphenyl-p-phenylenediamine
NMMANG-monomethyl-L-arginine
ROSreactive oxygen species
LDHlactate dehydrogenase